summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2016-08-27[bug] let the failures propagateRuben Pollan
In the case of wrong username/password the failure returned was 'Authentication hook did not return anything'. Right now the original failure is returned, and the producer of the failure can put a more meaningful message. - Resolves: #8399
2016-08-25[tests] pep8 fixes + add pep8 step to tox runKali Kaneko
2016-08-25[tests] move sample file to incoming test folderKali Kaneko
2016-08-25[tests] avoid pytest warningKali Kaneko
2016-08-25[tests] toxify leap.mailKali Kaneko
2016-08-25[tests] adapt smtp module testsKali Kaneko
2016-08-25[tests] adapt and fix outgoing module testsKali Kaneko
keymanager was hanging because it was trying to fetch a nonexistent key. therefore, fetch_remote flag has to be passed along.
2016-08-25[tests] fix initialization of incoming service testsKali Kaneko
otherwise, there was a very ugly bug in which the (imap) inbox kept a reference to the first instance of soledad used during a testing session. that made tests hang because, when that soledad instance is shutdown, the decryption pool is no longer running.
2016-08-25[tests] adapt imap testKali Kaneko
2016-08-25[tests] fix testKali Kaneko
2016-08-25[bug] fix import time mutable default paramKali Kaneko
2016-08-25[tests] adapt top level mail module testsKali Kaneko
2016-08-25[refactor] move common testing boilerplateKali Kaneko
2016-08-25[refactor] [bug] simplify and bugfix walk moduleKali Kaneko
some tests added too - Related: #7999
2016-08-01[refactor] deprecating u1db, using l2db instead.NavaL
To keep compatibility with soledad upgrades. It will namely cause version conflicts to not be properly handled otherwise.
2016-07-28[pkg] remove simplejson depdrebs
2016-07-12[test] add soledad client as a test requirementdrebs
2016-07-12[test] add gitlab-ci yaml filedrebs
2016-07-12[test] remove extra setUpEnv/tearDownEnvdrebs
The setUpEnv and tearDownEnv methods are configured to run on setUpClass/tearDownClass. Some testing systems as trial don't support that, and that seems to cause some confusion in some systems, causing double attempts to setup or teardown, and thus failing intermitently. This commit removes the extra call to setUpEnv and tearDownEnv on tests on this repository, as it now uses py.test to run tests, which supports setUpClass and tearDownClass.
2016-07-12[test] add tox.ini filedrebs
2016-07-12[test] move tests to root of repodrebs
2016-07-05Merge remote-tracking branch 'pixelated/check_requirements-latest' into developRuben Pollan
2016-07-05[feat] make ValidationLevels iterableRuben Pollan
Needed by bytmask cli. - Related: #7965
2016-07-05remove links to pixelatedChristoph Kluenter
See https://github.com/pixelated/puppet-pixelated/issues/49
2016-06-22[bug] the api_uri in Dicovery is called netlocRuben Pollan
2016-06-20[feat] check key document versions and fail if it's unknownRuben Pollan
- Closes: #8165
2016-06-17[bug] initialize OpenSSL context just once in leap.mailNavaL
Do not initialize the openssl context on each call to get mail payload phash. The openSSL backend should only be initialized once because it is activating the os random engine which in turn unregister and free current engine first. This is very tricky when operations are running in threads as it essentially momentarily unregister the openssl crypto callbacks that makes openssl thread safe. - Resolves: #8180 with the soledad PR #324
2016-06-17[feat] improve duplicated active documents fixupRuben Pollan
Mostly for private keys we want to keep the latest one in the keyring. It would be desirable to check with one is set in nicknym, but this will need a big refactor that I'm not sure if it's worth it just for that. - Related: #6658, #8059
2016-06-10[bug] Add KeyNotFound importZara Gebru
Two tests where failing, because - Related: #7987
2016-06-08[tests] keymanager._fetcher doesn't exist anymoreRuben Pollan
- Resolves: #8177
2016-06-08Merge branch 'pixelated/fix_sign_apple_mail' into developRuben Pollan
2016-06-07Merge branch 'pixelated/fix_test_decrypt_email' into developRuben Pollan
2016-06-06[pkg] bump version compat for keymanagerKali Kaneko
2016-06-01[test] test key generation (skipped)Ruben Pollan
2016-06-01fixup! [feat] remove the keytype supportRuben Pollan
2016-06-01[feat] remove request dependencyRuben Pollan
2016-06-01[feat] make get_all_keys aware of active addressesRuben Pollan
- Resolves: #8068
2016-06-01[refactor] move soledad document definitions to it's own fileRuben Pollan
2016-06-01[feat] remove keytype from build_key_from_dictRuben Pollan
2016-06-01[feat] remove the keytype supportRuben Pollan
- Resolves: #8031
2016-06-01[refactor] remove EncryptionKeyRuben Pollan
2016-06-01[refactor] move TempGPGWrapper to it's own fileRuben Pollan
2016-06-01[refactor] remove EncryptionSchemeRuben Pollan
2016-06-01Merge remote-tracking branch 'pixelated/7987_Invalid_JSON_data_from_key' ↵Ruben Pollan
into develop
2016-05-30[bug] fix invalid json warning when key fetch failsZara Gebru
Raise KeyNotFound error, when nicknym throws response 404, as it fails to find a key and add tests. - Resolves: #7987
2016-05-25[refactor] change the check to validate signature from Apple MailCaio Carrara
It changes the way that incoming service checks if a additional verification is needed to validate signature. The way before was checking by the type of signature object and calling the verify signature method if the type is different from OpenPGPKey. However it could be more readable if we check the type of decrypted message. If it's a multipart/signed message and not a plain/text we need to verify the signature because keymanager couldn't do it during the decryption process.
2016-05-25Merge remote-tracking branch 'pixelated/6200_add_support_for_binary_key' ↵Ruben Pollan
into develop - Resolves: #6200
2016-05-24[tests] add test to validate signature from apple mailCaio Carrara
This change adds test to validate signature of encrypted email created by apple mail. It's important to note that apple mail has a specific way to encrypt signed messages. First it sign the email and then encrypt the previous signed message. It was also added a message file with the expected data.
2016-05-24Issue #287: add tests for binary key and refactorZara Gebru
2016-05-19[feat] use netloc instead of an extra api_uri in DiscoveryRuben Pollan